Factors associated with nonparticipation and loss to follow-up in bereavement research using data from a cohort of parents with a loss in pregnancy or the neonatal period.

Abstract

Factors shaping participation in longitudinal cohorts may introduce bias, but there is a lack of research concerning participation patterns in cohorts of bereaved. Our aim was to study this in a longitudinal cohort of parents with a loss in pregnancy or during the neonatal period. Using data from the Danish national registers we identified all parents who experienced a pregnancy or neonatal loss in 2016-2018 and linked that with data on participants from the Life After the Loss cohort. Analyzing 725 parents, baseline participation was 25.7%, with 72.6% and 59.7% continuing at the seven- and 13-month follow-up. Partners and parents with lower education were underrepresented at baseline and at follow-up. Although not statistically significant, substantial underrepresentation of parents with mental health conditions and of parents who lived alone was observed at baseline. Although initial participation was low, after enrollment, most parents engaged in follow-up, minimizing the risk of attrition bias.